SDXL

anima_pencil-XL_clear

anima_pencil-XL_clear - AI Model封面圖片
上傳時間
2026年3月17日 清晨6:44
使用數
509
評論
好  (2)
許可權
  • 允許圖片生成與分享
  • 允許使用者下載此模型
  • 允許生成圖片用於商業用途

推薦 LoRA

描述

anima_pencil-XL_clearGuide (External site): English | JapaneseHugging FaceCompare to anima_pencil-XL-v5.0.0 (Original)anima_pencil-XL_clear_v1 | anima_pencil-XL-v5.0.0 (Original)A fine-tuned version of anima_pencil-XL-v5.0.0 model, designed to produce clearer and more vibrant outputs.Compared to the original, this model generates illustrations with higher contrast, richer colors, and more refined details.Recommended Settings (v1)GeneralCFG scale: 2-4Sampler: DPM++ 2M SDEscheduler: betaSteps: 15-25 (around 20)Quality Tags and Negative Prompts are not always necessary.Upscaleryour choice, examples:RealESRGAN_x4Plus Anime 6B (Commercial use)IllustrationJaNai_V1_ESRGAN (No-Commercial use)Detailerworks well.Merge RecipeUNET: Suppression of overly reactive layers (e.g., time_embed., label_embed.)VAE: Replacement with an anime-specific VAE (SDXL Anime VAE Dec-only B3)Tips (External site)Steps and Scheduler: English | JapaneseUpscalers: English | JapaneseCFG and Negative Prompt: English | JapaneseImproving Prompt Encoding in SDXL: English | JapaneseLicenseFair AI Public License 1.0-SD

留言

由 anima_pencil-XL_clear 製作的作品

1